sign in sign up
<<
Home , Benmoreite, Hancock Seamount, Honolulu Volcanics, Laysan, La Perouse Pinnacle, Macdonald seamount

VOLCANISM IN Chapter 1

- .-...... ,. THE HAWAIIAN-EMPEROR VOLCANIC CHAIN Part I Geologic Evolution

By David A. Clague and G. Brent Dalrymple

ABSTRACT chain, the near-fixity of the hot spot, the chemistry and timing of The Hawaiian-Emperor volcanic chain stretches nearly the eruptions from individual volcanoes, and the detailed geom­ 6,000 km across the North Pacific Ocean and consists of at least etry of volcanism. None of the geophysical hypotheses pro­ t 07 individual volcanoes with a total volume of about 1 million posed to date are fully satisfactory. However, the existence of km3• The chain is age progressive with still-active volcanoes at the Hawaiian ewell suggests that hot spots are indeed hot. In the southeast end and 80-75-Ma volcanoes at the northwest addition, both geophysical and geochemical hypotheses suggest end. The bend between the Hawaiian and .Emperor Chains that primitive undegassed material ascends beneath reflects a major change in Pacific plate motion at 43.1 ± 1.4 Ma Hawaii. Petrologic models suggest that this primitive material and probably was caused by collision of the Indian subcontinent reacts with the ocean to produce the compositional into Eurasia and the resulting reorganization of oceanic spread­ range of Hawaiian . ing centers and initiation of zones in the western Pacific. The volcanoes of the chain were erupted onto the floor of the Pacific Ocean without regard for the age or preexisting INTRODUCTION structure of the ocean . Hawaiian volcanoes erupt lava of distinct chemical com­ The ; the , hanks, and islands of positions during four major stages in their evolution and the Hawaiian Ridge; and the chain of Emperor Seamounts form an growth. The earliest stage is a submarine alkalic preahield array of shield volcanoes that stretches nearly 6,000 km across the stage, which is followed by the tholeiitic shield stage. The shield north Pacific Ocean (fig. 1.1). This unique geologic feature consists stage probably accounts for >95 percent of the volume of each . The shield stage is followed by an alkalic postshield of more than 107 individual volcanoes with a combined volume 3 stage during which a thin cap of alkalic and associated slightly greater than 1 million km (Bargar and Jackson, 1974). The differentiated lava covers the tholeiitic shield. After several chain is age progressive with still-active volcanoes at the southeast million years of , alkalic rejuvenated-stage lava erupts end whereas those at the northwest end have ages of about 75-80 from isolated vents. An individual volcano may become extinct Ma. The volcanic ridge is surrounded by a symmetrical depression, before the sequence is complete. The alkalic preshield stage is only known from recent study of Loihi . Lava from the Hawaiian Deep, as much as 0.7 km deeper than the adjacent later eruptive stages has been identified from numerous sub­ ocean floor (Hamilton, 1957). The Hawaiian Deep is in turn merged volcanoes located west of the principal Hawaiian surrounded by the broad Hawaiian Arch. Islands. At the southeast end of the chain lie the eight principal Volcanic propagation rates along the chain are 9.2 ± 0.3 em! Hawaiian Islands. Place names for the islands and seamounts in the yr for the Hawaiian Chain and 7.2 ± 1.1 cm/yr for the Emperor Chain. A best fit through all the age data for both chains gives chain are shown in figure 1.1 (see also table 1.2) The Island of 8.6±0.2 em/yr. Alkalic rejuvenated-stage lava erupts on an Hawaii includes the active volcanoes of , which erupted older shield during the formation of a new large in 1984, and Kilauea, which erupted in 1986. Loihi Seamount, 190±30 km to the east. The duration of the quiescent period located about 30 km off the southeast coast of Hawaii, is also active preceding eruption of rejuvenated-stage lava decreases system­ and considered to be an embryonic Hawaiian volcano (Malahoff, atically from 2.5 m.y. on to <0.4 m.y. at Haleakala, reflecting an increase in the rate of volcanic propagation during chapter 6; Moore and others, 1979, 1982). Hualalai Volcano on the last few million years. Rejuvenated-stage lava is generated Hawaii and Haleakala Volcano on have erupted in historical during the rapid change from subsidence to uplift as the vol­ times. Between Niihau and Kure Island only a few of the volcanoes canoes override a flexural arch created by loading the new rise above the sea as small volcanic islets and atolls. Beyond shield volcano on the ocean lithosphere. Kure the volcanoes are entirely submerged beneath the sea. Paleomagnetic data indicate that the Hawaiian hot spot has remained fixed during the last 40 m.y., but prior to that time the Approximately 3,450 km northwest of Kilauea, the Hawaiian hot spot was apparently located at a more northerly latitude. Chain bends sharply to the north and becomes the Emperor The most reliable data suggest about 70 of southward movement Seamounts, which continue northward another 2,300 km. of the hot spot between 65 and 40 Me. It is now clear that this remarkable feature was formed during The numerous hypotheses to explain the mechanism of the the past 70 m.y. or so as the Pacific lithospheric plate moved north hot spot fall into four types: propagating hypotheses, and then west relative to a melting anomaly, called the Hawaiian hot thermal or chemical convection hypotheses, shear melting hypotheses, and heat injection hypotheses. A successful spot, located in the asthenosphere. According to this hot-spot hypothesis must explain the propagation of volcanism along the hypothesis, a trail of volcanoes was formed and left on the ocean 5

u.s. Geological Survey Professional Paper 1350

... w, , 8 VOLCANISM IN HAWAII

140· 160· 1BO· 120·

BERING SEA NORTH AMERICA

400

20'

EXPLANATION Age of oceanic crust (Mal

.0-40 .106-120

.40-73 0120-150

R73-106 .>150

FIGURE 1.2.-Magnetic anomaliesand structure of ocean floor crust in North Pacific modified from Hilde and others (1976} Hawaiian-Emperor Chain (black) crosscuts preexisting fracture zones and Mesozoic magnetic-anomaly sequence.

and small flows of lava richer in SiOz and FeO (these are the the bulk of the shield is built of tholeiitic basalt, but during declinin hawaiite and that characterize the alkalic postshield activity the compositions revert to alkalic basalt. The alkali stage). S. Powers (1920) noted that eruptive centers of preshield stage, like the alkalic postshield stage, produces only smai basalt on , , , and Maui were active long after volumes of lava, probably totaling less than a few percent of th the main volcano became quiet; he appears to have been the first to volcano. associate nepheline basalt with late-stage eruptions following an We have omitted the main -collapse stage of Stearn erosional hiatus. (1966) from the eruption sequence because it can occur either durin: New insight into the preshield stage has come from recent the shield stage or near the beginning of the alkalic postshield stage studies of Loihi Seamount, a small located about The lava erupted may therefore be tholeiitic or alkalic basalt, or 0 30 km off the southeast coast of Hawaii. Its location, small size, both types. seismic activity, and fresh, glassy lava all indicate that Loihi is an active volcano and the youngest in the Hawaiian-Emperor Chain. GEOLOGY OF THE HAWAIIAN ISLANDS Some of the older lava samples recovered from Loihi Seamount are alkalic basalt and , whereas the youngest lava samples Descriptions of volcanoes and their eruptions were made b: recovered are tholeiitic and transitional basalt. This observation led nearly all the earliest visitors to the Hawaiian Islands. Description Moore and others (1982) to conclude that Loihi Seamount, and of particular note are those of William Ellis (1823), George, Lon perhaps all Hawaiian volcanoes, initially erupt alkalic basalt. Later, Byron (1826), Joseph Goodrich (1826, 1834), and Titus COal I. THE HAWAIIAN·EMPEROR VOLCANIC CHAIN PART I 9

TABLE 1. I.-Hawaiian eruptive producb 100 HAWAIIAN RIDGE

Eruptive stage Rock types Eruption Volume rate (percent)

Rejuvenated -- alkalic basalt Very low basanite nephe 1ini t e ! nepheline melilitite 0 0 0 ~ 0 N 0 N, e ~ Postshield --- alkalic basalt Low ~ 0 transitional basalt ~ s ~ ankaramite ~ 0 hawaiite > 0 mugearIte . g ~ § -c benmoreite , ,, "~ I phone lite Shield ------tholeiitic basalt High 95-98 20 tholeiitic basalt w picritic tholeiitic basalt o icelandite (rare) -c (rare) lalkalic basalt (1)

o 2000 4000 Preshield ---- basanite Low alkalic basalt DISTANCE FROM KILAUEA, IN KILOMETERS transitional basalt ltholeiitic basalt (7)

FIGURE 1.3.-Age of oceanic crust when overlying volcano formed, as a function lWright and Helz (chapter 23) suggest that the shield stage may of distance from Kilauea, for selected volcanoes in Hawaiian-Emperor Chain. include rare intercalated alkalic basalt and that the preshield stage includes tholeiitic basalt. We suspect that tholeiitic and alkalic Note offsets at fracture zones. Along Hawaiian Ridge both crust and volcanoes basalt occur intercalated during the transitions from preshield to increase in age 10 west so crustal age when volcanoes formed is roughly constant. shield stage and from shield stage to poatshield stage but that during On the other hand, Emperor Seamounts increase in age 10 north but crust the main shield stage only tholeiitic lava is erupted. decreases in age; thus age of crust when seamounts formed decreases from roughly 75 Ma at the bend to less than 40 Ma at Suiko Seamount.

milestone because it added detailed descriptions and chemical analyses of rocks from Hawaiian volcanoes other than Kilauea and Mauna Loa. (1840 and other letters until 1882). Many of their letters describing More detailed petrographic descriptions of lava from the the volcanoes were published in the American Journal of Science. islands were puhlished by S. Powers (1920). Soon afterward, Goodrich, in particular, provided detailed descriptions of the vol­ papers appeared by Washington (1923a, b, c) and Washington and canoes on the Island of Hawaii. None of the earliest descriptions Keyes (1926, 1928) with delailed accounls of the geology and however, included information about the mineralogy or of petrology, new high-quality chemical analyses of lava from Hawaii the lava. and Maui, and a classification of Hawaiian volcanic rocks. Palmer The Exploring Expedition visited Hawaii in (1927, 1936) added geologic descriptions and petrography of lava 1840-41. The commander of the expedition published a narrative from Kaula and Islands, both of which are cones of the (Wilkes, 1845) containing descriptions of caldera activity at Kilauea alkalic rejuvenated stage. Lehua Island is just one of several and new maps of both Kilauea and Mauna Loa . James D. rejuvenated-stage vents associated with Niihau, whereas Kaula Dana, the geologist of the expedition, published a detailed report on Island sits atop a completely submerged shield. the geology of the areas visited by the expedition (Dana, 1849). These early, mainly descriptive and reconnaissance studies This report contains descriptions of lava flows including their were superseded by detailed mapping of the islands beginning in the mineralogy and flow morphology, in addition to numerous other 1930's. H.T Stearns and his coworkers, in a remarkable series of observations on the active and inactive volcanoes that make up the bulletins published by the Hawaii Division of Hydrography, pub­ islands. Later reports by Dutton (1884), J.D. Dana (1887, 1888, lished geologic maps and descriptions of Oahu (Stearns and 1889), Green (1887), and Brigham (1909) added details on Vaksvik, 1935; Stearns, 1939, I940h), and eruptions and expanded the geologic observations to other islands. (Stearns. I940c), Maui (Stearns and Macdonald, 1942), Hawaii E.S. Dana (1877), Cross (1904), and Hitchcock (1911) (Stearns and Macdonald, 1946), Niihau (Stearns, 1947), and presented detailed petrographic descriptions of lava from the Molokai (Stearns and Macdonald, 1947). The bulletin on Keuai islands. Daly (1911) and Cross (1915) described the mineralogy by Macdonald and others (1960) completed the monumental map­ and petrology of Hawaiian lava flows at the time the Hawaiian ping job begun by Stearns; though Stearns did not coauthor the Volcano Ohservatory was established, and Jagger (1917) described report, he did much of the mapping and is an author of the map. activity in Halemaumau lava . The paper by Cross (1915) is a These maps and bulletins provide the geologic framework for all 10 VOLCANISM IN HAWAII

TABLE 1.2.-Ernptive stages represented on volcanoes of the Hawaiian-Emperor Chain (Volcano numbers from Bargar and Jackson (1974); numbers 0 and 65A added for consistency. Presence of slages: M, major unit; R, ran' or of small volume: X, present but c;lenl unkflown, A, known to be absent; -, unknown. (T), Iransitionallava probably erupted during late shield stage or caldera-collapse phase of that stage. For volcanoes from Kilauea through Nec::ker. data from detailed mapping and sampling; for remaining volcanoes, primarily from dredge and drill samples1

Eruptive Stages Volcano Preshield Shield postshield Rejuvenated Number Name (a1.kalie) (tholeiitic) (alkalic) (alkalic)

Hawaiian Islands

o Loihi ------­ M M 1 Kilauea ------M A A 2 Mauna Loa ------­ M A A 3 ------­ M M A 4 Hualalai ------M M A 5 ------M M A 6 East Maui ------­ M M M 7 Kahoolawe ------­ M R R 8 West Maui ------M R R 9 Lanai ------M A A 10 East Molokai ------­ M M R 11 West Molokai ------M R A 12 Koolau ------M A M 13 Waianae ------­ M M R 14 Kauai ------M R M 15 Niihau ------M R M 15A Kaula ------X X X

Northwestern Hawaiian Islands and Hawaiian Ridge

17 ------­ M 19 (Unnamed Seamount) X( T) 20 (Unnamed Seamount) --­ X X 21 (Unnamed Seamount) --­ X 23 Necker ------­ M X X 26 La Perouse Pinnacles ­ X 28 Brooks Bank ------­ X( T) X 29 St. Rogatien Bank ---­ X 30 ---­ X X 36 ------­ X 37 Northampton Bank ----­ X 39 pioneer Bank ------­ X 50 Pearl and Hermes ­ X 51 Ladd Bank ------­ X 52 Midway ------­ M X 53 Nero Bank ------­ X 57 (Unnamed Seamount) --­ X 63 (Unnamed Seamount) --- x 65 Calahan ------X X 65A Abbott ------X( T)

Emperor Seamounts

67 Daikakuji ------­ X 69 Yuryaku ------­ X X 72 Kimmei ------­ X 74 Koko (southern) -----­ X M 76 Koko (northwest) ----­ X 81 Ojin ------­ X X 83 Jingu ------­ X 86 Nintoku ------­ X 90 Suiko (southern) ----­ X 91 Suiko (central) -----­ M X 108 Meiji ------M I. THE HAWAIIAN·EMPEROR VOLCANiC CHAIN PART I II subsequent studies of the islands and can also be used to put many of GEOLOGY OF THE HAWAIIAN RIDGE the earlier descriptions into a broader geological context. A number of derivative publications include summaries of the geology of the The Northwestern Hawaiian Islands were the focus of all islands by Stearns (1946, 1966), an overview of the petrography of geologic investigations along the Hawaiian Ridge west of Kauai lava from the islands by Macdonald (1949), and a summary of the until oceanographic techniques were applied to the area in the geology of the Hawaiian Islands by Macdonald and others (1983). 1950's. Geological descriptions of the leeward islands include those The brief geologic summaries in appendix 1. I have largely been of S. Powers (1920) for Nihoa and Necker Islands, and Wash­ extracted from the above publications. Additional unpublished ington and Keyes (1926) and Palmer (1927) for Nihoa, Necker, observations by ourselves are included for Hualalai, East and West Gardner Pinnacles, and French Frigates Shoal (LaPerouse Pinna­ Molokai, Kooleu. Kauai, and Niihau. cles). These reports cite earlier sketchy descriptions. Macdonald The maps of Stearns and coworkers separate rejuvenated-stage (1949) reexamined Palmer's samples and added more detailed lava from earlier lava, but do not subdivide shield and postshield petrography. The petrology of the basaltic of Midway lava on the basis of chemical composition. The eruptive stages that Atoll is described from two drill cores by Macdonald (1969) and are known to occur in each of the volcanoes of the Hawaiian Islands Dalrymple and others (1974, 1977), whereas the geology of the site are summarized in table 1.2. Evidence for the alkalic preshield stage is detailed by Ladd and others (1967, 1969), Paleomagnetic data exists only at Loihi Seamount. If this stage is present in all Hawaiian on flows and dikes from N ihoa and Necker Islands are given by volcanoes, it is completely buried by later, shield-stage tholeiitic Doell (1972), whereas similar data from the Midway drill core are lava. The tholeiitic shield stage is known to form the major portion of given by Gromme and Vine (1972). the subaerial and, we assume, the submarine part of each volcano. Marine geologic investigations of the Hawaiian Ridge began On the main islands, only Hualalai Volcano and Kaula Island do with Hamilton's pioneering work in 1957. Much subsequent work not have subaerial exposures of tholeiitic lava. Alkalic lava of the has focused on the structure of the oceanic crust in the vicinity of the alkalic postshield stage occurs relatively late in the eruptive sequence chain, but few cruises have actually been conducted that dealt mainly and has not yet developed on Loihi Seamount or Kilauea and with the geology of the Hawaiian Ridge. In the early 1970's, Mauna Loa Volcanoes. To the northwest of there it occurs on all Scripps Institution of Oceanography and the Hawaii Institute of volcanoes except Lanai and Koolau, although the volumes present Geophysics conducted cruises to the Hawaiian Ridge. Samples on Kauai, Niihau, Kahoolawe, and West Molokai are small. Some collected by these cruises are described in Clague (1974a, 1974h) volcanoes have predominantly mugearite, whereas others have pre­ and Garcia, Grooms, and Naughton (in press). Subsequent cruises dominantly hawaiite; these are called Kohala type and Haleakala to the area by the Hawaii Institute of Geophysics and the U.S. type, respectively, by Macdonald and Katsura (1962). Wright and Geological Survey are cited in appendix 1. 1. Clague (in press) propose two additional types: a Hualalai type Lava samples recovered from the Hawaiian Ridge and with a bimodal trachyte-alkalic basalt lava distribution and a Emperor Seamounts are more difficult to assign to volcanic stages Koolau type with little or no alkalic postshield lava present. because they are recovered by dredging and drilling or are collected Hawaiian volcanoes commonly have summit calderas and from small islets and the field relations are usually unknown or only elongate curved zones from which much of the lava issues. poorly known. Tahle l .2 summarizes the available data from the Summit calderas exist on Loihi Seamount (Malahoff, chapter 6; Hawaiian Ridge. Based on the sequence and volumes of lava in the Malahoff and others, 1982), Kilauea, and Mauna Loa. Each of Hawaiian Islands, we have assumed that tholeiitic basalt always these calderas is connected to two prominent rift zones. Not all represents the shield stage and that strongly alkalic, SiOz-poor lava Hawaiian volcanoes, however, had a summit caldera. West Molokai represents the alkalic rejuvenated stage. Differentiated alkalic lava Volcano, in particular, shows no evidence of ever having had a has been assigned to the alkalic postshield stage. Some alkalic basalt caldera. Flat-lying lava ponded inside a caldera is not exposed on occurrences could be assigned to either the alkalic postshield or Hualalai, Mauna Kea, Kohala, or Niihau, but former calderas are rejuvenated stages; they have been assigned on the basis of trace­ inferred at those volcanoes from geophysical data (see Macdonald element signatures and mineral chemistry using criteria outlined in a and others, 1983). later section of this paper. No lava samples have been assigned to the The formation and structure of the rift zones have been alkalic preshield stage because we assume that the small volumes of examined in an elegant paper by Fiske and Jackson (1972~ who such early lava have been buried by the later voluminous tholeiitic concluded that the orientation of the rift zones reflects local gravita­ lava of the shield stage. tional stresses within the volcanoes. Isolated shields such as Kauai The samples recovered by dredging are probably not represen­ and West Molokai had nearly symmetrical stress fields represented tative of the lava forming the bulk of the individual seamounts, but by generally radial dikes and thus have only poorly defined rift instead represent the youngest lava types erupted on the volcanoes. zones. The rift zones of these isolated volcanoes tend to align parallel This natural sampling bias should result in an overrepresentation of to the orientation of the chain, suggesting the influence of a more alkalic lava from both the postshield and rejuvenated stages. In regional stress field that also controls the orientation of the chain. In addition, selection of recovered samples for further study introduces contrast, the rift zones of the other volcanoes tend to be aligned another bias because the freshest samples are commonly alkalic lava, parallel to the flanks of the preexisting shields against which they particularly hawaiite, mugearite, and trachyte. With these biases in abut. mind, it is still possible to note general trends along the entire chain. 12 VOLCANISM IN HAWAII

Tholeiitic basalt and picritic tholeiitic basalt, similar to those of samples are alkalic basalt, hawaiite, mugearite, and trachyte similar the shield stage of subaerial Hawaiian volcanoes, have been to lava erupted in the Hawaiian Islands, but lava from Koko recovered from II seamounts, banks, and islands in the Hawaiian Seamount includes anorthoclase trachyte and that are Ridge west of Kauai and Niihau (table 1.2; appendix 1.1). The interpreted to have erupted during the alkalic postshield stage abundance of tholeiitic basalt from the Hawaiian Ridge implies that (Clague, 1974a). Lava of the rejuvenated alkalic stage has not been these volcanoes are genetically related to the Hawaiian Islands and identified from any of the Emperor Seamounts. that the general sequence of Hawaiian volcanism, in which tholeiitic basalt forms a major portion of each volcano, has occurred along the SUBSIDENCE OF THE VOLCANOES entire Hawaiian Chain.

Charles Darwin (1837, 1842) was the first to suggest that GEOLOGY OF THE EMPEROR SEAMOUNTS coral atolls might grow on subsiding platforms and that drowned Little was known of the geology of the Emperor Seamounts atolls and certain deeply submerged banks with level tops could be until quite recently. The chain was recognized as the continuation of explained by subsidence. Hess (1946) recognized that flat-topped the Hawaiian Ridge by Bezrukov and Udintsev (1955), but not submarine peaks, which he named guyots, were drowned islands. until 1968 were the first samples recovered from Suiko Seamount He thought that they were volcanic, bare of and coral, (Ozima and others, 1970). These samples are dominantly, if not and had been planed off by erosion at . He attributed their completely, ice-rafted detritus. Subsequent studies included a cruise depth to rising sea level caused by deposition in the oceans. to the southern part of the chain by Scripps Institution of Menard and Dietz (1951) agreed with Hess that submergence was Oceanography in 1971 (ARIES Leg VII; Davies and others, primarily due to a rise in sea level, but they thought that local 1971, 1972), Deep Sea Drilling Project (DSDP) Site 192 on Meiji subsidence might also playa role. Hamilton (1956), in his classic Seamount (Creager and Scholl, 1973), DSDP Sites 308 and 309 study of the Mid-Pacific , which included a program of on Koko Seamount (Larson and others, 1975), a cruise by the dredging and coring, concluded that those (and other) guyots were Hawaii Institute of Geophysics (Dalrymple and Garcia, 1980), a formerly basaltic islands that had been wave and stream eroded and cruise by the U.S. Geological Survey in 1976 that surveyed the sites on which coral reefs subsequently grew. Their eventual sub­ for Leg 55 of DSDP (Dalrymple and others, 1980a), and Leg 55 mergence, he thought, was primarily caused by regional subsidence DSDP Sites 430, 431, 432, and 433 in the central part of the of the sea floor. It is now known that Darwin and Hamilton were chain (jackson and others, 1980). The Scripps Institution of basically correct about the steps leading to the formation of guyots, Oceanography cruise ARIES VII in 1971 and the Leg 55 DSDP and about the predominant role of subsidence in the process. cruise in 1977 were particularly successful, and most of our The Hawaiian-Emperor volcanic chain is an excellent example knowledge of the Emperor Seamounts is derived from these two of the gradual transformation of volcanic islands to guyots. From cruises. southeast to northwest there is a continuous progression from the The petrology of lava samples recovered by these two cruises is active volcanoes such as Mauna Loa and Kilauea through the described in Clague (1974a) and Kirkpatrick and others (1980), eroded remnants of Niihau, Nihoa, and Necker, through growing respectively. A detailed seismic interpretation of the caps atolls like French Frigates Shoal and Midway lslands, to deeply of many of the seamounts is given by Greene and others (1980), and submerged guyots like Ojin and Suiko. The progression can be overviews of the results of DSDP Leg 55 are given by Jackson and observed not only along the chain but within the stratigraphy of others (1980) and Clague (1981). individual seamounts. Drilling, dredging, and seismic observations Table 1.2 summarizes the available data on eruptive stages have shown conclusively that the atolls and guyots of the chain are represented by samples from the Emperor Seamounts, and details capped by carbonate deposits that overlie subaerial lava flows (see, are given in appendix 1.1 for individual volcanoes. We have for example, Ladd and others, 1967; Davies and others, 1971, assumed that tholeiitic basalt represents the shield stage and that 1972; Greene and others, 1980; Jackson and others, 1980). alkalic lava postdates the tholeiitic shield stage; only at Ojin and The subsidence of Hawaiian volcanoes with time results from Suiko Seamounts does drilling show that the alkalic lava overlies the thermal aging of the lithosphere and isostatic response to local tholeiitic flows. loading. The depth of the sea floor (and of volcanoes sitting upon it) Tholeiitic basalt and picritic tholeiitic basalt similar to those of increases away from spreading ridges because the lithosphere cools, the shield stage of subaerial Hawaiian volcanoes have been thickens, and subsides as it moves away from the source of heat recovered by drilling and dredging from six volcanic edifices in the beneath the ridge (Parsons and Sclater, 1977; Schroeder, 1984). Emperor Seamounts. The abundance of tholeiitic lava from the Detrick and Crough (1978) pointed out that the subsidence of many Emperor Seamounts is strong evidence that these volcanoes are islands and seamounts, including those along the Hawaiian­ genetically related to the Hawaiian Islands and Hawaiian Ridge. Emperor Chain, was far in excess of that which could be accounted Likewise, the general eruptive model for the Hawaiian Islands is for by this normal lithospheric aging or by lithospheric loading. apparently applicable to the Emperor Seamounts. They proposed that the lithosphere is thermally reset locally as it Alkalic postshield-stage lava has been recovered by dredging passes over a hot spot and that the excess subsidence is largely a and drilling from nine seamounts in the chain. In general these consequence of renewed lithospheric aging. I. THE HAWAIIAN-EMPEROR VOLCANIC CHAIN PART I 13

The Hawaiian-Emperor Chain rests on crust of Cretaceous GEOCHRONOLOGY AND PROPAGATION OF age (circa 120-80 Ma) for which the depth should be about VOLCANISM 5.5-5.9 km. The depth near Hawaii, however, is less than 4.5 km (fig. 1.4). The depth increases along the chain to about 5.3 km near EARLY WORK, LEGENDS AND DEGREE OF EROSION the Hawaiian-Emperor bend in a manner consistent with the thermal resetting hypothesis. Thus, the subsidence of Hawaiian volcanoes as According to Hawaiian legend, the goddess first inhab­ they move away from the hot spot is in part a function of their ited Kauai, but then moved southeastward island by island to distance from the hot spot, that is, of the reset thermal age of the Kilauea Volcano, where she now resides (Bryan, 1915~ The lithosphere beneath the chain. The volcanoes are passively riding reasoning behind this legend is unknown, but it was probably based away from the Hawaiian hot spot on cooling and thickening in large part on the relative appearance of age of the various lithosphere that is subsiding at about 0.02 nun/yr. volcanoes. Many centuries after this legend originated, J.D. Dana Superimposed on the effect of crustal aging is subsidence (1849) rendered the first scientific opinion confirming the general age caused by the immense and rapid loading of the lithosphere by the progression implied by the legend. growing volcanoes (Moore, chapter 2). This effect is local, but Dana was not only the first geologist to conclude that the order while the volcano is active the rate of subsidence caused by loading of extinction of Hawaiian volcanoes was approximately from north­ may exceed that from lithospheric aging by more than two orders of west to southeast, he also recognized that the Hawaiian Chain magnitude. Moore (1970) found, from a study of tide-gage records included the islets, atolls, and banks that stretch for some distance to in the Hawaiian Islands and on the west coast of North America, the northwest of Kauai. Dana saw no reason to think that the that Hilo on the Island of Hawaii has been subsiding at an absolute volcanoes of the chain did not originate simultaneously: "No facts rate of 4.8 mm/yr since 1946. Recent data on drowned coral reefs can be pointed to, which render it even probable that Hawaii is of near Kealakekua indicate an absolute subsidence rate for the more recent origin than Kenai" (Dana, 1849, p. 280} Their western side of Hawaii of 1.8 to 3 + mm/yr averaged over the past relative degree of erosion. however, provided ample evidence to 300,000 yr and also indicate that the rate may have accelerated indicate their order of extinction: "From Kauai to Mount Loa all during that time (Moore and Fornari, 1984). Moore's (1970) tide­ may thus have simultaneously commenced their ejections, and have gage data also show that absolute subsidence decreases systemati­ continued in operation during the same epoch till one after another cally away from the Island of Hawaii, with rates for Maui and Oahu became extinct. Now, the only burning summits out of the thirteen of 1.7 mm/yr and 0 mm/yr, respectively. Some of this decrease in which were once in action from Niihau to Hawaii, are those of Loa subsidence may be due to compensating uplift as the volcanoes are and Hualalai: we might say farther that these are all out of a number carried over the Hawaiian Arch, but an analysis of gravity data unknown, which stretched along for fifteen hundred , the length indicates that there is no appreciable viscous reaction to the sea­ of the whole range. This appears to be a correct view of the mount loads over time (Watts, 1978} Thus, it is probable that the Hawaiian Islands" (Dana, 1849, p. 280). Subsequent workers volcanoes are isostatically compensated within a few million years of agreed with Dana on the general order of extinction (for example, their birth, and that thermal aging of the lithosphere is the major Brigham, 1868; Dutton, 1884; Hillebrand, 1888; Hitchcock, cause of subsidence along the chain. 1911; Cross, 1915; Martin and Pierce, 1915; Wentworth, 1927; Hinds, 1931; Stearns, 1946), although the sequences they proposed invariably differed in detail (table 1.3) Of these various workers only Stearns (19461 who studied the Hawaiian Islands in more detail than any of his predecessors, had the sequence exactly correct as judged by present data. 4,----,--,--,------==--,------, The idea that the volcanoes of the Hawaiian Chain originated V> HAWAII MIDWAY I ••• , '"w simultaneously and only became extinct progressively seems to have tu ...... persisted until a few decades ago. Stearns (1946), for example, o mentions the lack of evidence to indicate when any of the Hawaiian "~ .. " 5 .... volcanoes began but shows all of the main shields except Hualalai ~ . and Kilauea erupting simultaneously at the end of the Pliocene I Ii: :....- 1- ••• (Stearns, 1946, p. 97, fig. 25). Two exceptions were Cross (1904) w ~iFZ- o - -MOFZ------__ and Wentworth (1927), who thought that the degree of erosion was 6L-_-l.______'__~______'__ _l__ _l__ _l__ __'__._J probably a function of when the volcanoes emerged above the sea as o 10 20 30 40 well as of the elapsed time since they ceased to erupt. Cross (1904, TIME SINCE REHEATING, IN MILLION YEARS p. 518) states: "It appears to me plausible to assume that the earliest FIGURE IA.-Minimum depth to sea-floor swell as a function of time since eruptions occurred at or near the western limit of this zone (the more reheating (or age of vokanoes along chain} Dashed line is predicted depth for than 1000 expanse of the island chain), and that in a general normal aging of lithosphere away from spreading ridge. Solid line is predicted depth for thermally reset lithosphere 45 km thick. MoFZ and MuFZ, Molokai way at least, the centers of activity have developed successively and Murray Fracture Zones, respectively. Modified from Detrick and Crough farther and farther to the east or southeast, until now the only active (1978) and Crough (1983). loci of eruption are those of Mauna Loa and Kilauea on the island of 14 VOLCANISM IN HAWAII

TABLE 1 3.-Earfy csnmcres of the order of extinction of the principal Hawaiian volcanoes [Criteria used are given beneath each source; volcanoes listed ill proposed order of extinction, oldest et tcp]

Dana ( 1849) Brigham (lsba) Dana (1888) Hillebrand (1888) Wentworth (1927) Hinds (1931) Stearns (1946 ) Erosion Erosion Erosion Floral divers ity Erosion Eros ion Erosion and stratigraphy

Kauai West Kauai, Kauai West Oahu, Kauai Koolau Waianae Kauai Waianae Niihau Waianae Molokai, East Oahu Kauai Koolau Waianae West Maui Waianae West MaUL Kcha l e , West Maui East Molokai Niihau Koolau Koolau East Kauai Kuhala Mauna K" West MaUL Kauai West Molokai Mauna K" West Molokai Koolau East Maui Mauna K" West Molokai East Molokai East Haui West Maui East Haui Hualalai Waianae East Molokai West Haui, Lanai Mauna Loa Kohala Mauna K" Mauna Loa, Kilauea East Maui Lanai Kahoolawe Koolau Hualalai Lanai West Maui, East Maui East Holokai Mauna Loa, Niihau, Kohala Kohala Mauna K" Kilauea West Holokai Kahoolawe Mauna K" Lanai, Kahoolawe Kahoolawe East Maui Hau l a l a i , Mauna East Maui Kohala Mauna K" Loa, Ki lauea Hualalai Hualalai, Mauna Hualalai Mauna Loa, Loa, Kilauea Mauna Loa, Ki Lauea Ki l a ue a

Hawaii. " He specifically noted the difference between his hypothesis have risen above the ocean long before, perhaps even in Mesozoic and that of Dana. time" (Hinds, 1931, p. 205), On the basis of geomorphic consid­ Estimates of the geologic ages of the Hawaiian volcanoes erations, Stearns (1946) thought that the volcanoes of the mam varied considerably among those early workers willing to hazard a Hawaiian Islands rose above sea level in the Tertiary. guess on the basis of the meager data then availahle. Dana (1849) thought it likely that the eruptions commenced as early as early Carboniferous or Silurian time; this estimate was based on the RADIOMETRIC AND AGES concept that the had cooled from a molten globe producing fissuring and volcanism, the apparent lack of post-Silurian volcanism The first radiometric ages for Hawaiian volcanoes were deter­ in the interior of the North American continent, and the presump­ mined by McDougall (1963), who measured ages of 2.8 to 3.6 Ma tion that the oceans would cool after the continents. Cross (1904) for his Middle and Upper Waianae Series on Oahu (all K-Ar ages speculated that the western part of the leeward islands formed in the have been converted to the new constants; Steiger and Jaeger, 1977). early part of the Tertiary. Wentworth (1925, 1927) attempted to He also reported an age of 8.6 Ma for what he called the Mauna quantify erosion rates for several of the islands and estimated the Kuwale Trachyte of the Lower Waianae Series(?), an age that later extinction ages of some of the volcanoes as follows: proved to be incorrect, probably because of excess in the analyzed (Funkhouser and others, 1968). In subsequent Lanai 0.15 Ma studies McDougall and Tarling (1963) and (primarily) McDougall Kohala 0.22 Ma (1964) reported K-Ar ages of lava from 7 of the principal Hawaiian Koolau 1.00 Ma volcanoes and concluded that the ages of the shield stages were Kauai 2.09 Ma approximately: On the basis of physiographic evidence, Wentworth doubted that Kauai 5.8-3.9 Ma any part of the Hawaiian group emerged above sea level before late Waianae 3.5-2.8 Ma Tertiary time. Hinds (1931), like Cross (1904), recognized that the Koolau 2.6-2.3 Ma atolls and banks of the leeward islands were the remnants of once­ West Molokai 1.8 Ma larger volcanoes: "The landscapes of the leeward group-the East Molokai 1.5-1.3 Ma volcanic stacks, the reef and calcareous sand islands rising West Maui 1.3-1.15 Ma from submarine platforms, and submerged platforms from which no East Maui 0.8 Ma islands rise, represent the final stages in the destruction of a volcanic Hawaii (all 5 volcanoes)